Feed_audio transcription

[Baavro]

I am unable to use feed_audio for transcription. It would be great if you could make a code snippet for the same.
i am sending data as
const audioBlob = new Blob(audioChunks, { type: 'audio/wav' });

  const reader = new FileReader();
  reader.readAsDataURL(audioBlob);

  reader.onloadend = () => {
    const audioBase64 = reader.result as string;

    const message = JSON.stringify({
      type: 'audio',
      spokenText: spokenText,
      content: audioBase64,
    });
    
    then I am feeding the content to feed_audio.

data_url = data.get("content")
_, base64_data = data_url.split(',', 1)
#audio_chunk = data.get("content")
audio_data = base64.b64decode(base64_data)
print("audio_file converted")
recorder.feed_audio(audio_data)
print("audio_file fed")

But I am unable to transcribe this

[KoljaB]

This example showcases the usage.
feed_audio expects chunks to be in raw 16-bit PCM mono format with samplerate 16000. Base64 is just a wrapper, it does not tell us about the internal format of the audio data. You need to ensure that the Base64-decoded data is converted into raw PCM format (if it isn't already) and matches the expected parameters (mono, 16 bit, 16kHz sample rate) before feeding it into the feed_audio method for transcription.